One of the goals of seismicity research is to find means to predict the hazard arising from seismicity. To achieve this it is necessary to link the mechanism of seismic event initiation to local rock failure. Damage simulation may provide this critical connection. The purpose of this paper is to demonstrate the potential of a numerical simulation technique that applies a statistical damage theory to seismicity study. The first part of this paper involves the relation between damage parameter and acoustic emissions caused by the failure of local elements of rock by considering the principles of both damage and acoustic emissions. A rationale for numerical simulation of acoustic emissions in rock failure is presented. In the second part, a numerical method, RFPA2D (Rock Failure Process Analysis), is used to model the evolution of progressive failure leading to collapse in uniaxial loading of brittle materials. The numerical compression tests show three typical patterns of acoustic emissions depending on the heterogeneity of the rock, which are in general agreement with the current understanding of the acoustic emissions. An example of simulating seismicities induced during the progressive failure of a pillar leading to final collapse is also given.
Rock deformation is accompanied by the formation of microfractures which emit elastic energy as acoustic emissions (AE). Since AE parameters such as event counts are related to damage accumulation, AE technique provides a unique perspective on the development of micro-fractures in rock subjected to applied load, and may be one of the best ways to monitor the damage during rock deformation and to predict the occurrence of a rock failure shock or instability. Lockner and Byerlee (1992), for example, draw upon laboratory results concerning the nucleation and growth of shear fractures in brittle rock. With AE monitoring, they have tracked the fault growth in granite sample, and have used their observations to develop a conceptual model that describes the fault nucleation based on the crack interaction. Cox et al. (1993) studied the microcrack formation and material softening in rock measured by monitoring acoustic emissions. While numerical simulations are popular and successful in describing the deformation behavior of rock before failure, there are few models that can simulate the progressive failure and induced seismicity (AE) in heterogeneous rock in a satisfactory manner. The majority of efforts has relied on elastic or plastic analyze to determine the potential for failure or yielding by comparing the calculated stresses with various strength criterion without considering the effects of heterogeneity on the stress redistribution and the failure process (Ferrero, et al, 1995; Kemeny, 1991; Plischke, 1991; Sterpi, et al, 1995; etc.). In this paper we simulate the progressive failure process and induced seismicities numerically using the Rock Failure Process Analysis code, RFPA20, developed at Northeastern University, P.R.China (Tang, 1995). RFPA2D allows to track the fracture process on a continuous basis during the loading process. By this analysis, the effect of heterogeneity on AE patterns can be considered and will be demonstrated on numerical compression tests.